Despite the success in elucidating the etiologies of several Mendelian diseases, defining the genetic architecture of common complex diseases, such as hypertension, a major cause of cardiovascular disease, remains a daunting challenge. We now have the technology to sift through large number of genetic variants to pinpoint those underlying specific diseases. However, for dissecting the genetic basis for complex diseases, there is currently no established framework for optimal study designs or analytic strategies. This application aims to develop a novel paradigm for studying complex diseases, in which a multi-stage design and a combination of analytic methods enable us to efficiently isolate genetic variants that influencing the risk of hypertension. Specifically, we propose to perform an admixture mapping study in African Americans using ancestry-informative SNP markers, followed by gene-based case-control association studies in the well-characterized cohorts recruited by the Family Blood Pressure Program (FBPP). We will also conduct further replication studies in different African-American cohorts and estimate population specific risks for the identified variants in European-American, Mexican American and Nigerian populations, respectively. The Specific Aims are: 1. Refine information on existing candidate regions on chromosomes 6 and 21 by admixture mapping. 2. Conduct association studies in the genes in refined regions. 3. Identify common variants in genes explaining the signals in admixture mapping. 4. Characterize effects of variants confirmed in Aim 3 in other population samples. Our prior research identified two new candidate regions (6q and 21 q) for hypertension (Zhu et al. 2005). Admixture mapping using highly informative SNPs (Aim 1) may achieve a greater power compared to conventional linkage study, and achieves a higher resolution compared to our previous admixture mapping analysis, which used a genome-wide microsatellite marker panel designed for linkage analysis. Overall, our multi-stage design is likely to achieve greater power and higher resolution compared to a single-stage design. [unreadable] [unreadable] [unreadable]